Bio
Parth Asawa is a PhD student at UC Berkeley advised by Professor Matei Zaharia and Professor Joey Gonzalez. Parth's research is on continual learning, studying how to enable models to stably learn from streams of experiences over time. His work focuses on sample-efficient learning and spans the stack of data, learning algorithms, architectures, and evaluation.
Session (1)
Additional Speaker Details
Using publicly available information we constructed an analysis to help you get a feel for this speaker before deciding to attend their session.
Asawa is a UC Berkeley CS PhD student (advised by Matei Zaharia and Joey Gonzalez) working on continual learning for LLM systems — how models can learn from streams of experience over time. He is first author on Continual Learning Bench, SIEVE, and Advisor Models, and ships the tooling behind that research, so his session offers a research-grounded, technically detailed view of how AI systems can actually improve from experience.
Speaking style
Recent talks (1)
GitHub
@pgasawa- pgasawa/continual-learning-bench★ 164Continual Learning Bench (CL-Bench): expert-validated benchmark measuring whether LLM-based agents genuinely improve from experience.
- pgasawa/origin★ 78Browser extension that uses LLMs to redefine the browser experience.
- pgasawa/BARE★ 41Base-Refine (BARE): leveraging base language models for few-shot synthetic data generation.
- pgasawa/SIEVE★ 11Sample-Efficient Parametric Learning from Natural Language.
- pgasawa/cs182-project★ 2Fine-tuning LLMs for style-specific text generation (course project).
Recent writing (4)
- Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environments · paper · Jun 2026
- SIEVE: Sample-Efficient Parametric Learning from Natural Language · paper · Feb 2026
- How to Train Your Advisor: Steering Black-Box LLMs with Advisor Models · paper · Oct 2025
- Unsafe AI or Consolidated Power: AI's False Dichotomy · blog · Jun 2026